anyone upgraded to 14.1 p6, what does it show in fetches like neofetch or fastfetch for you does it show up as p5 for you as well?

polyduekes Some updates only patch the userland and don't affect the kernel.

polyduekes see: freebsd-version -ku

k (kernel) / u (userland)

devnull: i see, thanks for the help

Hmm I have this weird thing with the p6 update. I updated with freebsd-update fetch / install and it now says there are no more updates. But uname still says p5?

Ooops sorry that question was just asked sorry

Thanks devnull

hi

Anyone else getting "fatal error: 'arm_neon.h' file not found" during buildworld on arm64?

I am having a problem with Python. It gives me an error: module 'importlib' has no attribute 'machinery'. Do you know what package I am missing?

what do you do to get that error? importlib is a core part of python

hi

please help, I need to install nvidia-driver 390 for an old nvidia 610, but even if I install it with pkg, the kernel complains about nvidia driver 550 installed...

Posterdati: pkg install x11/nvidia-driver-390

Did not work for you?

i recently noticed that transfers between disks connected to different controllers seem to be limited to like 40MB/s, while the disks are capable of 100+MB/s, and I indeed see that transfer speeds that high when data moves between disks on the same controller. is this considered normal?

this is an older machine (~2012), but it still seems weird.

phryk: no, it is not

Alver: thanks for confirming, will do some more benchmarking when i get around to it, but the mainboard will soon be retired, so i guess i'll mostly wait and see if it magically resolves itself with the new hardware. :)

Might be better indeed. Chasing ghosts of soon to be decommissioned systems is unlikely to be worth it

ye, it's served me long and well, only added two new disks that don't fit on the same controller anymore last year and gonna add two new ones soon. on the new hardware the last 2 will go onto an pcie-sata controller so that might bring its own issues, but it's AHCI-compatible and one PCIe 4.0 lane should still be plent for two platter disks. :)

More than plenty

Pauli1: I only get 800x600 graphics mode

Pauli1: this is an optimus graphics nvidia + intel

has anyone worked with the ports back: freshports.org/misc/ollama meta LLM? When I run the program under a debian bhyve it "works" but when i try to transition to a frebsd jail and this package.. it ALWAYS defaults to wanting a GPU.. where the deiban version does not.

Hello everyone. I keep getting a python error (importlib has no attribute machinery). Do you know if I need some certain package?

msamsing: you shouldn't need to. at least on my system "/usr/local/lib/python3.11/importlib/machinery.py" comes from the python311 package.

doesn't look like machinery needs a direct import (i.e. import importlib.machinery) either, so i'd look at what importlib.__file__ says, maybe it's not the actual importlib but some sorta shim?

while we're at python, i'm currently trying to execute python from a cronjob and i see that the command is executed, but don't get any further information – but the side effects that i expected don't happen. how do i debug this?

phryk: i have the machinery.py in the same path. trying to use gpg with offlineimap but that is where I get the error

no idea what offlineimap is, but gpg is a can of worms. least worst interface to it (at least like 7 years back when i last used it) was pretty-bad-protocol by isis lovecruft.

context: phryk.net/article/zomgwtftls/#gnupg---oh-eris-why

(site might intermittently go down, currently not completely done with deployment)

i would recommend sequoia, but last i checked it didn't have python bindings. maybe vishwin can weigh in on that, they took over maintainership of the port a while back.

phryk: I will have a look at that. The weird thing is that the same small peace of code works in a Linux system and an OpenBSD system?

yeah, had similar experiences when using the official gpgme python bindings back in the day. was a good part of the reason i ended up using pretty-bad-protocol.

my nvme nda1 decided to offline itself after lots of errors

but IMO gpg is just inherently bad because it doesn't have an actual API and everything just goes through the CLI.

I rebooted (so it reappeared), cleared, and let it resilver & scrubs fine

is there any sort of `nvmecontrol` thingy that would give me some sort of error details?

It's about 2 weeks left in warranty, so I can get it replaced if I can produce some plausible looking error allegations

dch: don't nvme drives support SMART?

i dont think in any useful style, but its a good idea, let me try ...

I was digging around with `nvmecontrol logpage -p1 -x nvme1` and similar, also not helpful

note for future nvme people, install `sysutils/nvme-cli` and get useful output

I have an issue with usbhidaction. I've configured it for volume and play/pause keys. I can use "mixer" to set volume but not "pactl". And I can't use "playerctl" for play/pause. I guess this is a dbus-issue but not sure how I can debug it.

With playerctl I get error like: "Error spawning command line ?dbus-launch --autolaunch... ..."

last Friday, i mentioned i couldn't get the NIC detected w/ freebsd or ghostbsd on my Dell inspiron 15.

well i tried nomadbsd on it, and it turns out there is a driver for it! rtw880

nomadbsd loaded the driver & turned wlan0 on by itself. so my real problem is getting it to see SSIDs from my router.

fixed my cronjob, redirecting output to file is still the most useful unix feature. turns out my program is expecting to be run from within its own directory so cd foo && <command> works :)

ZedHedTed: been a long time since i actively used wlan, but IIRC you don't have to see the SSIDs if you explicitly set the MAC of the access point in your wpa_supplicant.conf

phryk: i see. i'll have to try that. i thought i remmeber the filesystem being read-only on nomadbsd & live versions of freebsd, but i'll try editing that conf.

so anyone used lird to duplicate a infrared controller on freebsd?

anyone hit by the interlock exploit yet?

how does it even spread?

FreeBaSeDGirl: yeah we're still cleaning it up at work.

that one?

yeap

what's the attack vector?

freebsd

ober: Can you share more details?

well, i see two recent ssh vulns, maybe that is in use

the interlock thing is just ransomware

doubt, hrm

who knows the entry point. the devs have been so locked out of the prod systems they tend to add back doors to debug.

Well yeah, kind of important to figure that part out, if you don't know how they got in, it may happen again..

well we have people from a tla talking with our security team.

their hardened buntu image will save us...

har

hi

please help, how can I run gl software on freebsd 14.1 + intel/nvidia optimus?

thanks!

you're welcome!

Posterdati Whats is gl software?

OpenGL

Thanks, rtprio. I don't have experience with this.

Posterdati: what errors are you getting ?

Y'know, seeing as my void install is on zfs I suppose I ought to be able to chuck a freebsd install into the same pool

that sounds like it could be a jumbled mess

it sure does

on the one hand I consider the thinkpad's install "disposable" in that I don't store critical data on it, and on the other hand it took me quite a while to get it set up nicely and it'd be nice to dual boot to see if I can do the same with bsd without having to yeet the linux install currently taking up the whole disk

or at least, i gather pools aren't resizeable

right, the size of the pool is a function of the layout and drives in the pool

You can expand the pool

If the drives are bigger than current pool

mm. well, it's one drive, one pool for now

honestly it was a real error not leaving a bare partition for swap

no swap; play it fast and loose.

haven't had discord or firefox oom-killed so far